DENGESİZ VERİ SETLERİ İÇİN İKİ AŞAMALI DENGELEME STRATEJİSİ: ADASYN İLE ÖRNEKLEM ARTIRMA, SVM TABANLI ÖRNEKLEM AZALTMA

Öz

Bu çalışma, makine öğrenmesi alanında sıkça karşılaşılan dengesiz veri sorununu ele alarak, azınlık sınıf örneklerinin çoğunluk sınıf tarafından gölgede bırakıldığı durumlara odaklanmaktadır. Böyle bir dengesizlik, sağlık hizmetlerinden finansal sahtekârlık tespitine ve IoT tabanlı endüstriyel süreçlere kadar pek çok alanda model performansını ciddi biçimde zayıflatır. Sorunu gidermek için, azınlık sınıfını sentetik örneklerle zenginleştiren ADASYN yöntemi, SVM tabanlı uzaklık ölçümüyle belirlenen “en uzak” %10’luk çoğunluk örneklerinin çıkarılmasıyla birleştirilmiştir. Önerilen yaklaşım, SVM, RF, XGBoost ve KNN sınıflandırıcılarıyla on farklı veri seti üzerinde test edilmiştir. Bunlar arasında hem kalite kontrol verilerini hem de IoT sensör ölçümlerini içeren, gerçek üretim ortamından elde edilmiş 'Tekstil' veri seti de yer almaktadır. Özellikle iplik kopması gibi nadir ancak üretim açısından kritik olayları barındıran bu veri seti, yoğun dengesizlik nedeniyle standart yöntemlerde düşük başarı sergilemektedir. G-Ortalamalar metriğinde önemli iyileşmeler sunan yöntem, azınlık sınıfın daha başarılı tespitine katkıda bulunmuş ve on veri setinden beşinde en yüksek G-Ortalamalar değerini elde etmiştir.

Anahtar Kelimeler

• Sınıflandırma
• Dengesiz Veri
• Örneklem Artırma
• Örneklem Azaltma
• Tekstil

A Two-Stage Balancing Strategy for Imbalanced Datasets: Oversampling with ADASYN and Undersampling Based on SVM

Öz

This study tackles the frequently encountered problem of imbalanced datasets in machine learning, focusing on cases where minority-class examples are overshadowed by the majority class. Such imbalance significantly undermines model performance in diverse fields, including healthcare, fraud detection, and IoT-based industrial processes. To address this issue, we combine the ADASYN method—enriching the minority class with synthetic samples—with the removal of the “most distant” 10% of majority-class instances identified via an SVM-based distance measure. The proposed approach is tested with SVM, RF, XGBoost, and KNN classifiers on ten different datasets. Among these is the “Textile” dataset, which includes both quality control data and IoT sensor measurements and was collected from a real world production environment. Notably, this dataset includes rare yet critical events such as yarn breakage, which standard methods fail to detect effectively due to pronounced class imbalance. Our approach achieves considerable enhancements in the G-Mean metric, thereby improving the detection of minority cases and securing the highest G-Mean values on five out of ten datasets.

Anahtar Kelimeler

• Classification
• Imbalanced Data
• Oversampling
• Undersampling
• Textile

Kaynakça

1. Abdelkhalek, A. ve Mashaly, M., (2023) Addressing the class imbalance problem in network intrusion detection systems using data resampling and deep learning, The journal of Supercomputing 79, 10611–10644. doi:10.1007/s11227-023-05073-x
2. Almarshdi, R., Nassef, L., Fadel, E. ve Alowidi, N. (2023) Hybrid deep learning based attack detection for imbalanced data classification, Intelligent Automation & Soft Computing 35. doi:10.32604/iasc.2023.026799
3. Altunkaynak, A., Başakın, E.E. ve Kartal, E. (2020) Dalgacik k-en yakin komşuluk yöntemi ile hava kirliliği tahmini, Uludağ Üniversitesi Mühendislik Fakültesi Dergisi 25, 1547– 1556. doi:10.17482/uumfd.809938
4. Amirshahi, B. ve Lahmiri, S. (2024) Bankruptcy prediction using optimal ensemble models under balanced and imbalanced data, Expert Systems, 41, e13599. doi:10.1111/exsy.13599
5. Avizheh, M., Dadpasand, M., Dehnavi, E. ve Keshavarzi, H. (2023) Application of machine-learning algorithms to predict calving difficulty in Holstein dairy cattle, Animal Production Science, 63, 1095–1104. doi: 10.1071/AN22461
6. Aymaz, S. (2025) Unlocking the power of optimized data balancing ratios: a new frontier in tackling imbalanced datasets, The Journal of Supercomputing, 81, 443. doi:10.1007/s11227-025-06919-2
7. Ercire, M. ve Ünsal, A. (2021) Kısa Süreli Güç Kalitesi Bozulmalarının Dalgacık Analizi ve Rastgele Orman Yöntemi ile Sınıflandırılması, Uludağ Üniversitesi Mühendislik Fakültesi Dergisi, 26, 903–920. doi:10.17482/uumfd.976342
8. Ferreira, G., Das, S., Coelho, G., Silva, R. R., Goswami, S., Pereira, R. N., Pereira, L., Fortunato, E., Martins, R. ve Nandy, S (2025) Energy harvesting and movement tracking by polypyrrole functionalized textile for wearable IoT applications, Journal of Energy Chemistry, 102, 230–242. doi: 10.1016/j.jechem.2024.10.028

9. Galán-Cuenca, A., Gallego, A. J., Saval-Calvo, M. ve Pertusa, A. (2024) Few-shot learning for COVID-19 chest X-ray classification with imbalanced data: An inter vs. intra domain study. Pattern Analysis and Applications, 27, 69. doi: 10.1007/s10044-024-01285-w
10. García-Gil, D., García, S., Xiong, N. ve Herrera, F. (2024) Smart data driven decision trees ensemble methodology for imbalanced big data, Cognitive Computation, 16, 1572–1588. doi: 10.1007/s12559-024-10295-z
11. Guo, J., Wu, H., Chen, X. ve Lin, W. (2024) Adaptive SV-Borderline SMOTE-SVM algorithm for imbalanced data classification, Applied Soft Computing, 150, 110986. doi: 10.1016/j.asoc.2023.110986
12. He, H., Bai, Y., Garcia, E. A. ve Li, S. (2008) ADASYN: Adaptive synthetic sampling approach for imbalanced learning, In Proceedings of the IEEE International Joint Conference on Neural Networks,1322–1328. doi: 10.1109/IJCNN.2008.4633969
13. He, H. ve Garcia, E. A. (2009) Learning from imbalanced data. IEEE Transactions on Knowledge and Data Engineering, 21, 1263–1284. doi: 10.1109/TKDE.2008.239
14. Hussein, H. I., Anwar, S. A. ve Ahmad, M. I. (2023) Imbalanced data classification using SVM based on improved simulated annealing featuring synthetic data generation and reduction, Computers, Materials & Continua, 75, 547–564. doi: 10.32604/cmc.2023.036025
15. Lappage, J. (2005) End breaks in the spinning and weaving of weavable singles yarns: Part 2: End breaks in weaving, Textile Research Journal, 75, 512–517. doi: 10.1177/0040517505053869
16. López, V., Fernández, A., García, S., Palade, V. ve Herrera, F. (2013) An insight into classification with imbalanced data, Information Sciences, 250, 113–141. doi: 10.1016/j.ins.2013.07.007
17. Markwald, M. ve Demidova, E. (2024) REFUEL: rule extraction for imbalanced neural node classification, Machine Learning, 113, 6227–6246. doi: 10.1007/s10994-024-06569-0
18. Martikkala, A., Mayanti, B., Helo, P., Lobov, A. ve Ituarte, I. F. (2023) Smart textile waste collection system – dynamic route optimization with IoT, Journal of Environmental Management, 335, 117548. doi: 10.1016/j.jenvman.2023.117548
19. Mebrate, M., Gessesse, N. ve Zinabu, N. (2022) Effect of loom tension on mechanical properties of plain woven cotton fabric, Journal of Natural Fibers, 19, 1443–1448. doi: 10.1080/15440478.2020.1776663
20. Militino, A. F., Goyena, H., Pérez-Goya, U. ve Ugarte, M. D. (2024) Logistic regression versus XGBoost for detecting burned areas using satellite images, Environmental and Ecological Statistics, 31, 57–77. doi: 10.1007/s10651-023-00590-7
21. Morgan, H., Elgendy, A., Said, A., Hashem, M., Li, W., Maharjan, S. ve El-Askary, H. (2024) Enhanced lithological mapping in arid crystalline regions using explainable AI and multi-spectral remote sensing data, Computers & Geosciences, 193, 105738. doi: 10.1016/j.cageo.2024.105738
22. Mu, X. ve Zhao, B. (2025) DCS-SOCP-SVM: A novel integrated sampling and classification algorithm for imbalanced datasets, Computers, Materials & Continua, 83, 2143–2159. doi: 10.32604/cmc.2025.060739
23. My, B. T. ve Ta, B. Q. (2023) An interpretable decision tree ensemble model for imbalanced credit scoring datasets, Journal of Intelligent & Fuzzy Systems, 45, 10853–10864. doi: 10.3233/JIFS-230825
24. Özdet, B. ve İçer, S. (2022) AKCİĞER BİLGİSAYARLI TOMOGRAFİ GÖRÜNTÜLERİNDE GÖRÜNTÜ İŞLEME UYGULAMALARI İLE TÜMÖRLERİNİN TESPİT EDİLMESİ, Uludağ Üniversitesi Mühendislik Fakültesi Dergisi, 27, 135–150. doi:10.17482/uumfd.947619
25. Peng, C. ve Yuan, X. (2021) A study on detection of roving tension and fine control of yarn breakage, Industria Textila, 72, 250–255. doi:10.35530/IT.072.03.1757
26. Pérez Tárraga, J., Castillo-Cara, M., Arias-Antúnez, E. ve Dujovne, D. (2025) Frost forecasting through machine learning algorithms, Earth Science Informatics, 18, 183. doi:10.1007/s12145-025-01710-6
27. Stanosheck, J.A., Castell-Perez, M.E., Moreira, R.G., King, M.D. ve Castillo, A. (2024) Oversampling methods for machine learning model data training to improve model capabilities to predict the presence of Escherichia coli MG1655 in spinach wash water, Journal of Food Science, 89, 150–173. doi:10.1111/1750-3841.16850
28. Tolentino-Zondervan, F. ve DiVito, L. (2024) Sustainability performance of Dutch firms and the role of digitalization: The case of textile and apparel industry, Journal of Cleaner Production, 459, 142573. doi:10.1016/j.jclepro.2024.142573
29. Wang, A.X., Chukova, S.S. ve Nguyen, B.P. (2023a) Ensemble k-nearest neighbors based on centroid displacement, Information Sciences, 629, 313–323. doi:10.1016/j.ins.2023.02.004
30. Wang, X., Ren, H., Ren, J., Song, W., Qiao, Y., Ren, Z., Zhao, Y., Linghu, L., Cui, Y., Zhao, Z. ve diğ. (2023) Machine learning-enabled risk prediction of chronic obstructive pulmonary disease with unbalanced data, Computer Methods and Programs in Biomedicine, 230, 107340. doi:10.1016/j.cmpb.2023.107340
31. Wang, Z. ve Liu, Q. (2023) Imbalanced data classification method based on LSSASMOTE, IEEE Access, 11, 32252–32260. doi:10.1109/ACCESS.2023.3262460
32. Weiss, G.M. (2004) Mining with rarity: a unifying framework, ACM SIGKDD Explorations Newsletter, 6, 7–19. doi: 10.1145/1007730.1007734
33. Wibbeke, J., Rohjans, S. ve Rauh, A. (2025) Quantification of data imbalance, Expert Systems, 42, e13840. doi: 10.1111/exsy.13840
34. Wu, L.j., Li, X., Yuan, J.d. ve Wang, S.j. (2023) Real-time prediction of tunnel face conditions using XGBoost random forest algorithm, Frontiers of Structural and Civil Engineering, 17, 1777–1795. doi: 10.1007/s11709-023-0044-4
35. Yao, G., Guo, J. ve Zhou, Y. (2005) Predicting the warp breakage rate in weaving by neural network techniques, Textile Research Journal, 75, 274–278. doi:10.1177/004051750507500
36. Yilmaz Eroglu, D. ve Pir, M.S. (2024) Hybrid oversampling and undersampling method (HOUM) via safe-level SMOTE and support vector machine, Applied Sciences, 14, 10438. doi: 10.3390/app142210438
37. Zhou, T., Gao, X., Sun, X. ve Han, L. (2024) Split difference weighting: An enhanced decision tree approach for imbalanced classification, International Journal of Computers Communications & Control, 19. doi: 10.15837/ijccc.2024.6.6702